It’s sad to see how far off the ledge For All Mankind has fallen. Like 99% of the show is just people walking around in corridors, and the one time they actually do want to show the space exploration part of titan it looks so cheap and low budget it’s sad. At this point why even stretch this to 6 seasons? Just end it already with the last episode, or give it a hell of a budget to do something more interesting than walk around cardboard corridors!!
Macro
Thanks to the weather our local town fair which is always a big deal has been a complete washout. I feel bad for all the vendors out there but then my teen and her friends who braved the rain and still went came back eating nothing or doing anything at the fair because as she said “everything is way too expensive!” after the pandemic our town is the only one that hosts fairs anymore while all the surrounding cities who used to throw these fairs have all shut down, so I hope this is not the death knell for our town too.
🍿 Dhurandhar: The Revenge (2026) – ★★★★☆
I actually kinda enjoyed this more than part 1. Solid 3 and half stars, but an extra half star because I sat through the entire 4 hour movie in one go when I thought it would take us several days to get through. I didn’t find the jingoism in this movie as problematic as the first one though. You can debate plot devices but it at least didn’t feel like forced upon. Superb soundtrack like the first one, and I do think the raw version on Netflix with all the gaali galouch makes it more real lol. All things said, movie really could have been tighter towards the end – there was no reason why this movie stretched to 4 hours.
What happens when rupee invariably crosses 100 to a dollar? Tech companies will prefer to keep it that way for their businesses, but overall imports would suffer to a large extent. It’s a horrible death spiral I assume though I am not an economist so I don’t really understand the implications much except “oh it can’t be good for anyone except MNCs”
Signed up the teen for another year in swim team but she was clearly not interested or enthusiastic this year. It took a lot of convincing from us for her to signup again. It’s absolutely true that teenage girls start dropping out of sports around this age, but we want to keep pushing her to continue till end of high school. I know where her inflection point happened – it was when she got injured in November last year which caused her to lose a spot in junior varsity team at school. It’s never been the same after that for her. Of course the coaches aren’t that great either as the team has had a lot of turnover.
The continuous glucose monitor is both a blessing and a curse. It’s lovely to be able to see what spikes your sugar in real time but for a health data nut like me I just can’t stop looking at it way too much. Ugh. at this rate I might not even enjoy a glass of water. I need to remind myself thst it’s ok to eat…. rich food once in a while, bhaad mei jaaye readings!
As I posted earlier in week I have been pushing Gemini long term memory in terms of health data. Today it finally broke apart. Look I am smart enough to know when to trust AI and when not to, but the fact that AI was actively trying to convince me that I am wrong, and then when proved wrong, refused to apologize and acknowledge the mistake was gyrating. Eventually I told it that I am not here to trust ai over a doctor, and it gave me a quick sorry and then tried to change the subject lol. Yeah, LLMs have fucking long long long way to go before they should be taken seriously beyond short term help. Will post in detail about the interaction later.
Now a days if a work meeting happens after core working hours due to east coast west coast time difference I just ask the organizer to transcribe the meeting. Then I use AI to summarize the meet for salient take aways. So much more optimal use of time and I don’t have sit through and a hour of ramble. Of course this works best when I am not the primary participant of the meeting, but those I try to schedule earlier in day anyway.
My LinkedIn is filled with people giving examples of Claude projects and readme.md files on how to set up Claude in an optimized way to use less tokens and be cheaper. And I am like why isn’t Claude doing all this itself? Why do I have to tell it what’s the best what to run agents and on which models. Do it yourself na? Surely it has to be coming?
Good grief I scolded Claude for incorrect market analysis based on my spreadsheet I provided it (it was wildly wrong) and now as if to penance for its gone off in the hinterlands for the last 15 minutes writing complicated code to read and analyze the excel file, redoing it over and over again coz I don’t the fuck know what’s it’s trying to do. Like Jesus dude, all you had to fucking do was take the table in 5th sheet and drop it in!